Editors’ Vox is a blog from AGU’s Publications Department.

Flooding is one of the most impactful natural disasters, disrupting the lives of many and causing tremendous costs worldwide each year. While scientists have gained a greater understanding of riverine and coastal floods over the past few decades, our modeling skills still have major limitations, which prevent accurate and efficient forecasting.

A recent study in Reviews of Geophysics explores the main challenges in riverine and coastal flood modeling and pathways for improvement in the future. We asked a few of the co-authors to provide an overview of the impacts and drivers of floods, describe what recent scientific advances have been made, and outline what questions remain.

Flooding is a well-known natural hazard. Is it possible to quantify the scale and cost of flooding worldwide?

Yes, we use a wide range of physical and statistical models to quantify the scale and cost of flooding worldwide. However, it is important to note that these quantifications are subject to a significant uncertainty, which can impact the accuracy and reliability of our results. Our primary goal is to enhance our modeling capabilities, reduce uncertainties, and deliver more dependable results for the benefit of the public, as well as emergency services and risk managers.

It is important to highlight that our review paper offers guidance on enhancing the quantification of flood scale, which typically involves generating flood extent and depth maps for various regions. However, quantifying the cost of flooding requires an additional level of analysis to translate the flood scale into tangible damages. The modeling techniques employed to estimate the damages introduce a new source of uncertainty. Therefore, we recommend a separate study that examines the potential challenges in flood vulnerability assessment, specifically focusing on the conversion of flood maps to potential damage.

How is climate change expected to affect flood risk? 

Climate change has shown to significantly impact precipitation patterns and contribute to increased inland and coastal floods via sea level rise and increased storm activities, both of which are key factors influencing the occurrence of riverine and coastal floods, respectively.

Global warming amplifies the intensity and frequency of rainfall events, consequently augmenting the risk of flash floods and urban flooding.

Global warming, a primary consequence of climate change, entails a long-term elevation in Earth’s average surface temperature. This temperature rise can have significant implications, such as earlier snowmelt in snow-dominated basins, which can contribute to more runoff and increased flooding during the springtime. Furthermore, larger moisture holding capacity of atmosphere is expected in a warmer climate. Numerous studies have demonstrated that global warming amplifies the intensity and frequency of rainfall events, consequently augmenting the risk of flash floods and urban flooding.

The shift in precipitation patterns and melting snow due to global warming affect the terrestrial flows and water levels and increase the frequency and magnitude of riverine flooding. In coastal regions, the rising temperature contributes to water expansion and the melting of glaciers and ice caps, leading to an escalation in sea levels. This, besides the possibility of increased storm activity, results in a higher likelihood of more frequent and intense coastal flooding events.

Why have many flood management strategies been unsuccessful? 

An important contributing factor to the shortcomings of flood management strategies is the incomplete understanding of nonlinear and complex climatic, hydrological, and hydrodynamic processes involved in flooding. The significant level of uncertainty associated with these processes poses challenges in accurately assessing flood hazards and risks in vulnerable regions. Consequently, the availability of suitable tools, methods, and technologies for comprehensive flood characterization and modeling is limited.

Improving our skills in flood characterization and modeling is paramount to developing more effective flood management strategies.

Improving our skills in flood characterization and modeling is paramount to developing more effective flood management strategies. However, it is crucial to acknowledge that other factors, such as lack of community engagement and awareness, insufficient financial support, and fragmented governance structures with conflicting interests among different entities, also contribute to the failure of flood management strategies. These factors extend beyond the scope of our review paper.

How does your review paper differ from previous studies on flooding?

Our review stands out from previous studies due to its comprehensive nature, encompassing a wide range of flood-related aspects. Prior works have predominantly concentrated on specific aspects of flooding. For instance, some review papers have exclusively examined flood forecasting and hydrological processes, while others have focused on flood inundation models and hydrodynamic processes. In contrast, our review encompasses an up-to-date analysis of the literature including both hydrologic and hydrodynamic processes.

Moreover, several crucial topics, such as flood inundation forecasting, causative mechanisms of floods, and coastal flooding, have not been extensively reviewed in the past. We endeavored to consolidate all pertinent subjects that contribute to the enhancement of flood characterization and modeling skills within a single, comprehensive review paper.

In simple terms, what are “causative mechanisms” of flooding and why do we need to better understand their role? 

Causative mechanisms of floods encompass a range of climatic, meteorological, hydrologic, and hydrodynamic processes that influence when, how long, and to what extent flooding occurs. These mechanisms can involve elements such as intense rainfall, snowmelt, saturated soil conditions, dam failures, coastal storm surges, or a combination thereof. Understanding these causative mechanisms is crucial as it enables us to incorporate all relevant factors into our flood models, providing a more comprehensive understanding of flood dynamics.

Gaining insight into the causative mechanisms of floods empowers us to make informed decisions to reduce the risk and consequences associated with flooding.

By comprehending the causative mechanisms, we enhance our flood modeling skills, leading to improved accuracy in predicting flood events. This, in turn, supports effective flood risk reduction strategies. By considering the various factors involved in flood generation, we can better assess the potential impacts, plan appropriate mitigation measures, and develop more robust flood management approaches. Ultimately, gaining insight into the causative mechanisms of floods empowers us to make informed decisions to reduce the risk and consequences associated with flooding.

How do the primary drivers of riverine and coastal floods differ? 

Riverine floods primarily occur when rivers and streams exceed their capacity due to factors such as heavy rainfall, significant snowmelt, or a combination of these events. This overflow of water leads to the inundation of surrounding areas and the potential for widespread flooding. To model riverine floods, inland precipitation is used as the main driver, and hydrological processes within the river systems are simulated.

Coastal floods, however, are primarily derived from tidal fluctuations, storm surges, waves and sea-level rise. Despite different spatiotemporal scales of these processes, when combined, they yield in large total water level that pose flood threat to the coastal communities. On a decadal time scale, processes including ocean warming and tectonic activities derive relative sea level rise that can significantly deviate from the global average of 3 millimeters/year. Along the East and Gulf coasts of the United States, for example, the local rates of sea level rise can be 3 to 4 times higher than the global average.

At higher frequencies, meteorological processes that cause atmospheric pressure fluctuations and wind  (i.e. seasonal variability or tropical cyclones) contribute to a higher coastal sea level than the typical tidal level that coastal communities expect in the absence of meteorological drivers. The combination of astronomical tides, storm surge and wind waves yield in elevated water level at the coast that leaves the communities nearby inundated and pushes large amounts of water landward.

Schematic of flood-generating processes from a hydrological (a), hydrodynamic (b), and biophysical processes (c) perspective. Credit: Jafarzadegan et al. [2023], Figures 1 and 2

What have been some recent advances in modeling and forecasting flooding? 

In general, the models used to forecast flooding have significantly improved including in the following ways:

  • improvement in the forecast of precipitation, sea level rise and storm surge used as the main input to riverine and coastal flood models 
  • advances in hydrological and hydrodynamic model structures allow for a more accurate representation of physical processes involved in flooding
  • advances in the integration of the weather forecast, hydrologic and hydrodynamic models within data assimilation frameworks using a wide range of satellite observations, radar data, and ground-based measurements
  • the access to fine-scale topographic data (e.g. 1 meter resolution LIDAR
    data) as input to hydrodynamic models allows more accurate simulation of
    flood dynamics 
  • advances in high-performance computing resources facilitates parallel simulation and ensemble-based flood modeling at larger scales and finer spatial and temporal resolutions.

Furthermore, our current operational flood forecasting systems can integrate numerical weather prediction models with hydrological and hydrodynamic modeling processes. This integrated framework offers a more comprehensive and synergistic approach to flood forecasting.

What are the main limitations and challenges of current flood modeling? What are the potential pathways to overcome these challenges?

In the concluding section of our review paper, we have presented an extensive compilation of limitations and challenges, and pathways in flood characterization and modeling. Here, we summarize some of the key points outlined:

  1. Incorporating additional factors into the analysis of flood-generating processes: Our current analyses primarily focus on hydroclimatic and hydrologic perspectives. However, other factors, such as topographic details of the study area, river morphology, and human interventions play significant roles in flood dynamics and should be integrated into our analyses.
  2. Limited use of Earth System Modeling framework for flood forecasting: Traditional operational flood forecast systems treat individual components separately. To improve accuracy and comprehensiveness, a transition is needed towards using earth system modeling frameworks. These frameworks enable the modeling of interactions between the atmosphere, oceans, land biosphere, and human activities within a unified framework, providing a more holistic understanding of flood processes.
  3. Limited access to a comprehensive dataset for flood modeling and validation: Access to fine-scale and accurate channel bathymetry data as well as flood defense locations is limited globally. Additionally, access to a comprehensive archive of high-quality flood extent observation maps related to historical floods is needed. This will facilitate rigorous and comprehensive testing of models, improving their reliability.
  4. The lack of comprehensive data for characterization of various aspects of coastal flooding. In areas subject to compound coastal flooding, the lack of spatiotemporal coverage of overlapping records to detect and characterize the interdependencies of compound flood drivers is a big challenge. Relatively short records of forcing data and observational coastal flooding data limits the opportunities to characterize nonstationarity in coastal flooding due to factors like sea level rise. Additionally, modeling wave spectrum at regional and global scales suffers from the limited range of data and environmental conditions considered for parameterization of the wave energy generation, propagation, and dissipation mechanisms.

—Keighobad Jafarzadegan (kjafarzadegan@ua.edu; 0000-0001-5840-5551), University of Alabama, USA; Hamed Moftakhari (0000-0003-3170-8653), University of Alabama, USA; and Hamid Moradkhani (0000-0002-2889-999X), University of Alabama, USA

Editor’s Note: It is the policy of AGU Publications to invite the authors of articles published in Reviews of Geophysics to write a summary for Eos Editors’ Vox.

Citation: Jafarzadegan, K., H. Moftakhari, and H. Moradkhani (2023), Unlocking the secrets of floods: breakthroughs in riverine and coastal modeling, Eos, 104, https://doi.org/10.1029/2023EO235017. Published on 5 July 2023.